DOCSLIB.ORG

Arxiv:1710.07394V2 [Cs.CL] 22 May 2018 Elucidate Public Opinion Surrounding Polarizing Quires Minimal Human Supervision and Can Be Eas- Events

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Arxiv:1710.07394V2 [Cs.CL] 22 May 2018 Elucidate Public Opinion Surrounding Polarizing Quires Minimal Human Supervision and Can Be Eas- Events Recognizing Explicit and Implicit Hate Speech Using a Weakly Supervised Two-path Bootstrapping Approach Lei Gao Alexis Kuppersmith Ruihong Huang Texas A&M University Stanford University Texas A&M University [email protected] [email protected] [email protected] Abstract contents) and tends to transform rapidly following a new trigger event. Our pilot annotation experi- In the wake of a polarizing election, social ment with 5,000 randomly selected tweets shows media is laden with hateful content. To that around 0.6% (31 tweets) of tweets are hate- address various limitations of supervised ful. The mass-scale (Yahoo! Finance online com- hate speech classification methods includ- ments) hate speech annotation effort from Yahoo! ing corpus bias and huge cost of annota- (Nobata et al., 2016) revealed that only 5.9% of tion, we propose a weakly supervised two- online comments contained hate speech. There- path bootstrapping approach for an online fore, large amounts of online texts need to be an- hate speech detection model leveraging notated to adequately identify hate speech. In re- large-scale unlabeled data. This system cent studies (Waseem and Hovy, 2016; Kwok and significantly outperforms hate speech de- Wang, 2013), the data selection methods and an- tection systems that are trained in a super- notations are often biased towards a specific type vised manner using manually annotated of hate speech or hate speech generated in cer- data. Applying this model on a large quan- tain scenarios in order to increase the ratio of hate tity of tweets collected before, after, and speech content in the annotated data sets, which on election day reveals motivations and however made the resulting annotations too dis- patterns of inflammatory language. torted to reflect the true distribution of hate speech. Furthermore, inflammatory language changes dra- 1 Introduction matically following new hate “trigger” events, Following a turbulent election season, 2016’s dig- which will significantly devalue annotated data. ital footprint is awash with hate speech. Apart To address the various limitations of super- from censorship, the goals of enabling computers vised hate speech detection methods, we present to understand inflammatory language are many. a weakly supervised two-path bootstrapping ap- Sensing increased proliferation of hate speech can proach for online hate speech detection that re- arXiv:1710.07394v2 [cs.CL] 22 May 2018 elucidate public opinion surrounding polarizing quires minimal human supervision and can be eas- events. Identifying hateful declarations can bolster ily retrained and adapted to capture new types of security in revealing individuals harboring mali- inflammatory language. Our two-path bootstrap- cious intentions towards specific groups. ping architecture consists of two learning compo- Recent studies on supervised methods for on- nents, an explicit slur term learner and a neural net line hate speech detection (Waseem and Hovy, classifier (LSTMs (Hochreiter and Schmidhuber, 2016; Nobata et al., 2016) have relied on manu- 1997)), that can capture both explicit and implicit ally annotated datasets, which are not only costly phrasings of online hate speech. to create but also likely to be insufficient to ob- Specifically, our bootstrapping system starts tain wide-coverage hate speech detection systems. with automatically labeled online hateful content This is mainly because online hate speech is rela- that are identified by matching a large collection tively infrequent (among large amounts of online of unlabeled online content with several hateful slur terms. Then two learning components will be two-path bootstrapping system is able to jointly initiated simultaneously. A slur term learner will identify many more hate speech texts (214,997 v.s learn additional hateful slur terms from the auto- 52,958 v.s 112,535) with a significantly higher F- matically identified hateful content. Meanwhile, a score (48.9% v.s 19.7% v.s 26.1%), when com- neural net classifier will be trained using the au- pared to the bootstrapping systems with only the tomatically labeled hateful content as positive in- slur term learner and only the neural net classi- stances and randomly sampled online content as fier. In addition, the evaluation shows that the negative instances. Next, both string matching two-path bootstrapping system identifies 4.4 times with the newly learned slur terms and the trained more hateful texts than hate speech detection sys- neural net classifier will be used to recognize new tems that are trained using manually annotated hateful content from the large unlabeled collec- data in a supervised manner. tion of online contents. Then the newly identi- fied hateful content by each of the two learning 2 Related Work components will be used to augment the initially identified hateful content, which will be used to Previous studies on hate speech recognition learn more slur terms and retrain the classifier. The mostly used supervised approaches. Due to the whole process iterates. sparsity of hate speech overall in reality, the data The design of the two-path bootstrapping sys- selection methods and annotations are often bi- tem is mainly motivated to capture both explicit ased towards a specific type of hate speech or and implicit inflammatory language. Explicit hate speech generated in certain scenarios. For hate speech is easily identifiable by recognizing a instance, Razavi et al.(2010) conducted their clearly hateful word or phrase. For example: experiments on 1525 annotated sentences from a company’s log file and a certain newsgroup. (1) Don’t talk to me from an anonymous account Warner and Hirschberg(2012) labeled around you faggot coward, whither up and die. 9000 human labeled paragraphs from Yahoo!’s (2) And that’s the kind of people who support news group post and American Jewish Congress’s Trump! Subhumans! website, and the labeling is restricted to anti- In contrast, implicit hate speech employs circum- Semitic hate speech. Sood et al.(2012) studied locution, metaphor, or stereotypes to convey ha- use of profanity on a dataset of 6,500 labeled com- tred of a particular group, in which hatefulness can ments from Yahoo! Buzz. Kwok and Wang(2013) be captured by understanding its overall composi- built a balanced corpus of 24582 tweets consist- tional meanings, For example: ing of anti-black and non-anti black tweets. The tweets were manually selected from Twitter ac- (3) Hillary’s welfare army doesn’t really want counts that were believed to be racist based upon jobs. They want more freebies. their reactions to anti-Obama articles. Burnap and (4) Affirmative action means we get affirmatively Williams(2014) collected hateful tweets related second rate doctors and other professionals. to the murder of Drummer Lee Rigby in 2013. Furthermore, our learning architecture has a fla- Waseem and Hovy(2016) collected tweets using vor of co-training (Blum and Mitchell, 1998) in hateful slurs, specific hashtags as well as suspi- maintaining two learning components that concen- cious user IDs. Consequently, all of the 1,972 trate on different properties of inflammatory lan- racist tweets are by 9 users, and the majority of guage. By modeling distinct aspects of online hate sexist tweets are related to an Australian TV show. speech, such a learning system is better equipped Djuric et al.(2015) is the first to study hate to combat semantic drift, which often occurs in speech using a large-scale annotated data set. self-learning where the learned model drifts away They have annotated 951,736 online comments from the esteemed track. Moreover, training two from Yahoo!Finance, with 56,280 comments la- complementary models simultaneously and utiliz- beled as hateful. Nobata et al.(2016) followed ing both models to identify hate speech of different Djuric et al.(2015)’s work. In addition to the Ya- properties in each iteration of the learning process hoo!Finance annotated comments, they also an- is important to maintain the learning momentum notated 1,390,774 comments from Yahoo!News. and to generate models with wide coverage. In- Comments in both data sets were randomly sam- deed, our experimental results have shown that the pled from their corresponding websites with a fo- cus on comments by users who were reported to with slur term seeds. Tweets that contain one of have posted hateful comments. We instead aim to the seed slur terms are labeled as hateful. detect hate speech w.r.t. its real distribution, using The two-path bootstrapping system consists a weakly supervised method that does not rely on of two learning components, an explicit slur large amounts of annotations. term learner and a neural net classifier (LSTMs The commonly used classification methods in (Hochreiter and Schmidhuber, 1997)), that can previous studies are logistic regression and Naive capture both explicit and implicit descriptions of Bayes classifiers. Djuric et al.(2015) and No- online hate speech. Using the initial seed slur term bata et al.(2016) applied neural network models labeled hateful tweets, the two learning compo- for training word embeddings, which were further nents will be initiated simultaneously. The slur used as features in a logistic regression model for term learner will continue to learn additional hate- classification. We will instead train a neural net ful slur terms. Meanwhile, the neural net classi- classifier (Kim, 2014; Lai et al., 2015; Zhou et al., fier will be trained using the automatically labeled 2015) in a weakly supervised manner in order to hateful tweets as positive instances and randomly capture implicit and compositional hate speech ex- sampled tweets as negative instances. Next, both pressions. the newly learned slur terms and the trained neu- Xiang et al.(2012) is related to our research ral net classifier will be used to identify new hate- because they also used a bootstrapping method ful content from the unlabeled large collection of to discover offensive language from a large-scale tweets.
Load more

Recommended publications
  • Race and Membership in American History: the Eugenics Movement
    Race and Membership in American History: The Eugenics Movement Facing History and Ourselves National Foundation, Inc. Brookline, Massachusetts Eugenicstextfinal.qxp 11/6/2006 10:05 AM Page 2 For permission to reproduce the following photographs, posters, and charts in this book, grateful acknowledgement is made to the following: Cover: “Mixed Types of Uncivilized Peoples” from Truman State University. (Image #1028 from Cold Spring Harbor Eugenics Archive, http://www.eugenics archive.org/eugenics/). Fitter Family Contest winners, Kansas State Fair, from American Philosophical Society (image #94 at http://www.amphilsoc.org/ library/guides/eugenics.htm). Ellis Island image from the Library of Congress. Petrus Camper’s illustration of “facial angles” from The Works of the Late Professor Camper by Thomas Cogan, M.D., London: Dilly, 1794. Inside: p. 45: The Works of the Late Professor Camper by Thomas Cogan, M.D., London: Dilly, 1794. 51: “Observations on the Size of the Brain in Various Races and Families of Man” by Samuel Morton. Proceedings of the Academy of Natural Sciences, vol. 4, 1849. 74: The American Philosophical Society. 77: Heredity in Relation to Eugenics, Charles Davenport. New York: Henry Holt &Co., 1911. 99: Special Collections and Preservation Division, Chicago Public Library. 116: The Missouri Historical Society. 119: The Daughters of Edward Darley Boit, 1882; John Singer Sargent, American (1856-1925). Oil on canvas; 87 3/8 x 87 5/8 in. (221.9 x 222.6 cm.). Gift of Mary Louisa Boit, Julia Overing Boit, Jane Hubbard Boit, and Florence D. Boit in memory of their father, Edward Darley Boit, 19.124.
    [Show full text]
  • The N-Word : Comprehending the Complexity of Stratification in American Community Settings Anne V
    Union College Union | Digital Works Honors Theses Student Work 6-2009 The N-word : comprehending the complexity of stratification in American community settings Anne V. Benfield Union College - Schenectady, NY Follow this and additional works at: https://digitalworks.union.edu/theses Part of the Race and Ethnicity Commons, Sociology of Culture Commons, and the Urban Studies and Planning Commons Recommended Citation Benfield, Anne V., "The -wN ord : comprehending the complexity of stratification in American community settings" (2009). Honors Theses. 1433. https://digitalworks.union.edu/theses/1433 This Open Access is brought to you for free and open access by the Student Work at Union | Digital Works. It has been accepted for inclusion in Honors Theses by an authorized administrator of Union | Digital Works. For more information, please contact [email protected]. The N-Word: Comprehending the Complexity of Stratification in American Community Settings By Anne V. Benfield * * * * * * * * * Submitted in partial fulfillment of the requirements for Honors in the Department of Sociology UNION COLLEGE June, 2009 Table of Contents Abstract 3 Introduction 4 Chapter One: Literature Review Etymology 7 Early Uses 8 Fluidity in the Twentieth Century 11 The Commercialization of Nigger 12 The Millennium 15 Race as a Determinant 17 Gender Binary 19 Class Stratification and the Talented Tenth 23 Generational Difference 25 Chapter Two: Methodology Sociological Theories 29 W.E.B DuBois’ “Double-Consciousness” 34 Qualitative Research Instrument: Focus Groups 38 Chapter Three: Results and Discussion Demographics 42 Generational Difference 43 Class Stratification and the Talented Tenth 47 Gender Binary 51 Race as a Determinant 55 The Ambiguity of Nigger vs.
    [Show full text]
  • The Semantics and Pragmatics of Slurs and Thick Terms Bianca Cepollaro
    The Semantics and Pragmatics of slurs and thick terms Bianca Cepollaro To cite this version: Bianca Cepollaro. The Semantics and Pragmatics of slurs and thick terms. Philosophy. Uni- versité Paris sciences et lettres; Scuola normale superiore (Pise, Italie), 2017. English. NNT : 2017PSLEE003. tel-01508856 HAL Id: tel-01508856 https://tel.archives-ouvertes.fr/tel-01508856 Submitted on 14 Apr 2017 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. THÈSE DE DOCTORAT de l’Université de recherche Paris Sciences et Lettres PSL Research University Préparée dans le cadre d’une cotutelle entre Scuola Normale Superiore, Pisa et École Normale Supérieure, Paris La sémantique et la pragmatique des termes d’offense et des termes éthiques épais Ecole doctorale n°540 ÉCOLE TRANSDISCIPLINAIRE LETTRES/SCIENCES Spécialité Philosophie COMPOSITION DU JURY : Mme. JESHION Robin University of South California, Rapporteur M. VÄYRYNEN Pekka University of Leeds, Rapporteur Mme. BIANCHI Claudia Soutenue par Bianca Università Vita-Salute San Raffaele, Membre du jury CEPOLLARO Le 20 janvier 2017h Mme. SBISÀ Marina Università degli Studi di Trieste, Membre du jury Dirigée par Pier Marco BERTINETTO et Isidora STOJANOVIC The semantics and pragmatics of slurs and thick terms Bianca Cepollaro Abstract In this thesis I develop a uniform account of slurs and thick terms in terms of presuppositions.
    [Show full text]
  • Unconscious Racism Revisited: Reflections on the Impact and Origins of "The Id, the Ego, and Equal Protection"
    Georgetown University Law Center Scholarship @ GEORGETOWN LAW 2008 Unconscious Racism Revisited: Reflections on the Impact and Origins of "The Id, the Ego, and Equal Protection" Charles R. Lawrence III Georgetown University Law Center, [email protected] This paper can be downloaded free of charge from: https://scholarship.law.georgetown.edu/facpub/339 40 Conn. L. Rev. 931-978 (2008) This open-access article is brought to you by the Georgetown Law Library. Posted with permission of the author. Follow this and additional works at: https://scholarship.law.georgetown.edu/facpub Part of the Civil Rights and Discrimination Commons, and the Constitutional Law Commons GEORGETOWN LAW Faculty Publications April 2010 Unconscious Racism Revisited: Reflections on the Impact and Origins of "The Id, the Ego, and Equal Protection" 40 Conn. L. Rev. 931-978 (2008) Charles R. Lawrence III Professor of Law Georgetown University Law Center [email protected] This paper can be downloaded without charge from: Scholarly Commons: http://scholarship.law.georgetown.edu/facpub/339/ Posted with permission of the author CONNECTICUT LAW REVIEW VOLUME 40 MAY 2008 NUMBER 4 Article Unconscious Racism Revisited: Reflections on the Impact and Origins of “The Id, the Ego, and Equal Protection” CHARLES LAWRENCE III Twenty years ago, Professor Charles Lawrence wrote “The Id, The Ego, and Equal Protection: Reckoning With Unconscious Racism.” This article is considered a foundational document of Critical Race Theory and is one of the most influential and widely cited law review articles. The article argued that the purposeful intent requirement found in Supreme Court equal protection doctrine and in the Court’s interpretation of antidiscrimination laws disserved the value of equal citizenship expressed in those laws because many forms of racial bias are unconscious.
    [Show full text]
  • On the Power of Slurring Words and Derogatory Gestures
    Charged Expressions: On the Power of Slurring Words and Derogatory Gestures Ralph DiFranco, Ph.D. University of Connecticut, 2016 Slurs have the striking power to promulgate prejudice. Standard semantic and pragmatic theories fail to explain how this works. Drawing on embodied cognition research, I show that metaphorical slurs, descriptive slurs, and slurs that imitate their targets are effective means of transmitting prejudice because they are vehicles for prompting hearers to form mental images that depict targets in unflattering ways or to simulate experiential states such as negative feelings for targets. However, slurs are a heterogeneous group, and there may be no one mechanism by which slurs harm their targets. Some perpetrate a visceral kind of harm – they shock and offend hearers – while others sully hearers with objectionable imagery. Thus, a pluralistic account is needed. Although recent philosophical work on pejoratives has focused exclusively on words, derogation is a broader phenomenon that often constitutively involves various forms of non- verbal communication. This dissertation leads the way into uncharted territory by offering an account of the rhetorical power of iconic derogatory gestures and other non-verbal pejoratives that derogate by virtue of some iconic resemblance to their targets. Like many slurs, iconic derogatory gestures are designed to sully recipients with objectionable imagery. I also address ethical issues concerning the use of pejoratives. For instance, I show that the use of slurs for a powerful majority
    [Show full text]
  • A Critical Discourse Analysis of Interracial Marriages in Television Sitcoms
    University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange Doctoral Dissertations Graduate School 5-2014 Drawing the Primetime Color Line: A Critical Discourse Analysis of Interracial Marriages in Television Sitcoms Jodi Lynn Rightler-McDaniels University of Tennessee, Knoxville, [email protected] Follow this and additional works at: https://trace.tennessee.edu/utk_graddiss Part of the Critical and Cultural Studies Commons, Gender, Race, Sexuality, and Ethnicity in Communication Commons, and the Race, Ethnicity and Post-Colonial Studies Commons Recommended Citation Rightler-McDaniels, Jodi Lynn, "Drawing the Primetime Color Line: A Critical Discourse Analysis of Interracial Marriages in Television Sitcoms. " PhD diss., University of Tennessee, 2014. https://trace.tennessee.edu/utk_graddiss/2725 This Dissertation is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council: I am submitting herewith a dissertation written by Jodi Lynn Rightler-McDaniels entitled "Drawing the Primetime Color Line: A Critical Discourse Analysis of Interracial Marriages in Television Sitcoms." I have examined the final electronic copy of this dissertation for form and content and recommend that it be accepted in partial fulfillment of the equirr ements for the degree of Doctor of Philosophy, with a major in Communication and Information. Catherine A. Luther, Major Professor We have read this dissertation and recommend its acceptance: Elizabeth Hendrickson, Lori Amber Roessner, Barbara Thayer-Bacon Accepted for the Council: Carolyn R.
    [Show full text]
  • Learning "The N-Word"
    wrote some of the most nuanced and interesting essays I’ve read in a long time. Four of them are reprinted here. h LEARNING “THE N-WORD” Adam Cherry Before this course, and before reading “Teaching The N- Word” by Emily Bernard, I was of the mindset that words are words. Their meanings are arbitrarily selected, and so lan­ guage should be fluid. Therefore, a word like “nigger” can be erased or assigned a new, more positive meaning. However, my simplistic view has been complicated by the use of this word in our course texts and Bernard’s essay which struggles to work out its role in language. In her article, Bernard often contradicts herself when try­ ing to determine what role she wants the “n-word” to play in her life. She ultimately can’t come to one conclusion, perhaps because of her personal investment in the topic. The best way to summarize what Bernard learns is that “nigger” means a lot of different things to different people. Embedded in American culture, the word’s impact cannot be studied as one unified American experience. Bernard tells her students that she doesn’t hold a special place of pain for the “n-word,” only to turn around and admit she is lying. Her motivation for doing so seems to be that an honest admission could squelch classroom debate. In reality, Bernard sees some truth in her students’ fears that saying the word out loud could annihilate her. The “n-word” has deep seeds of influence in her life. It reminds her of the past strug­ gles of her race and creates anxiety in her own life.
    [Show full text]
  • Demographic Word Embeddings for Racism Detection on Twitter
    Demographic Word Embeddings for Racism Detection on Twitter Mohammed Hasanuzzaman Gael¨ Dias Andy Way ADAPT Centre GREYC UMR 6072 ADAPT Centre Dublin City University Universite´ de Caen Normandie Dublin City University Ireland France Ireland Abstract depression, anxiety, and emotional stress as well as negative physical health outcomes such as high Most social media platforms grant users blood pressure and low infant birth weight. freedom of speech by allowing them to As the information contained in social media freely express their thoughts, beliefs, and often reflects the real-world experiences of their opinions. Although this represents in- users, there is an increased expectation that the credible and unique communication op- norms of society will also apply in social media portunities, it also presents important chal- settings. As such, there is an increasing demand lenges. Online racism is such an exam- for social media platforms to empower users with ple. In this study, we present a super- the tools to report offensive and hateful content vised learning strategy to detect racist lan- (Oboler and Connelly, 2014). guage on Twitter based on word embed- Hateful content can be defined as ”speech or ex- ding that incorporate demographic (Age, pression which is capable of instilling or inciting Gender, and Location) information. Our hatred of, or prejudice towards, a person or group methodology achieves reasonable classi- of people on a specified ground, including race, fication accuracy over a gold standard nationality, ethnicity, country of origin, ethno- dataset (F1=76.3%) and significantly im- religious identity, religion, sexuality, gender iden- proves over the classification performance tity or gender” (Gelber and Stone, 2007).
    [Show full text]
  • The Rise and Fall of the Hillbilly Music Genre, a History, 1922-1939. Ryan Carlson Bernard East Tennessee State University
    East Tennessee State University Digital Commons @ East Tennessee State University Electronic Theses and Dissertations Student Works 12-2007 The Rise and Fall of the Hillbilly Music Genre, A History, 1922-1939. Ryan Carlson Bernard East Tennessee State University Follow this and additional works at: https://dc.etsu.edu/etd Part of the Musicology Commons Recommended Citation Bernard, Ryan Carlson, "The Rise and Fall of the Hillbilly Music Genre, A History, 1922-1939." (2007). Electronic Theses and Dissertations. Paper 2059. https://dc.etsu.edu/etd/2059 This Thesis - Open Access is brought to you for free and open access by the Student Works at Digital Commons @ East Tennessee State University. It has been accepted for inclusion in Electronic Theses and Dissertations by an authorized administrator of Digital Commons @ East Tennessee State University. For more information, please contact [email protected]. The Rise and Fall of the Hillbilly Music Genre: A History, 1922-1939. ___________________ A thesis presented to the faculty of the Department of Liberal Studies East Tennessee State University In partial fulfillment of the requirements for the degree Masters of Arts in Liberal Studies ___________________ by Ryan Carlson Bernard December, 2006 ___________________ Dr. Richard Blaustein, Chair Dr. Ted Olson Dr. Kevin O’Donnell Keywords: Hillbilly, Music, Stereotype, Genre, Phonograph, Radio ABSTRACT The Rise and Fall of the Hillbilly Music Genre: A History, 1922-1939 by Ryan Carlson Bernard This research will examine the rise in popularity of the hillbilly music genre as it relates to the early part of the twentieth century as well as its decline with the arrival of the western hero, the cowboy.
    [Show full text]
  • The Social Life of Slurs
    The Social Life of Slurs Geoff Nunberg School of Information, UC Berkeley Jan. 22, 2016 To appear in Daniel Fogal, Daniel Harris, and Matt Moss (eds.) (2017): New Work on Speech Acts (Oxford, UK: Oxford University Press). Chaque mot a son histoire. —Jules Gilliéron A Philological Caution The Emergence of Slurs We wear two hats when we talk about slurs, as engaged citizens and as scholars of language. The words had very little theoretical interest for philosophy or linguistic semantics before they took on a symbolic role in the culture wars that broke out in and around the academy in the 1980s.1 But once scholars’ attention was drawn to the topic, they began to discern connections to familiar problems in meta-ethics, semantics, and the philosophy of language. The apparent dual nature of the words—they seem both to describe and to evaluate or express— seemed to make them an excellent test bed for investigations of non-truth-conditional aspects of meaning, of certain types of moral language, of Fregean “coloring,” and of hybrid or “thick” terms, among other things. There are some writers who take slurs purely as a topical jumping-off point for addressing those issues and don’t make any explicit effort to bring their discussions back to the social questions that drew scholars’ attention to the words in the first place. But most seem to feel that their research ought to have some significance beyond the confines of the common room. That double perspective can leave us a little wall-eyed, as we try to track slurs as both a social and linguistic phenomenon.
    [Show full text]
  • Juke Joints, Rum Shops, and Honky-Tonks, the Politics Of
    Draft of a chapter in a forthcoming anthology, please do not cite “Juke Joints, Rum Shops, and Honky-Tonks: Black Migrant Labor and Leisure in Caribbean Guatemala and Honduras, 1898–1922” Frederick Douglass Opie Babson College In 1922, Eugene Cunningham and his companion, both white travelers from the United States, entered a combination bar and grocery store in Zacapa on the Caribbean coast of Guatemala. The store was operated by a white American expatriate. An African American customer, also an expatriate, invited the travelers to have a drink with him. When they ignored this request, the man ordered the two white men to join him, after which a violent altercation ensued. In the course of his travels, Cunningham also heard of a white bar operator who warned a racist Irish American expatriate that if he continued to call blacks “nigger,” “his end would come suddenly from a knife between the ribs.”1 What do anecdotes like these—anecdotes in which white men of European ancestry are threatened by black Americans in Guatemala—tell us about the relationship between alcohol and culture, class, and race on the Caribbean Coast of Guatemala in the late nineteenth and early twentieth centuries? Black immigrants went to Guatemala in the hope of finding better economic opportunities than were available to them in their home Draft of a chapter in a forthcoming anthology, please do not cite region. The labor market in lowland Guatemala was good because Guatemalan nationals refused jobs in malaria-infested regions where railroad track was laid. Black immigrants were willing to go there and work long hours for higher wages.
    [Show full text]
  • Racial Slur Database - by SLUR (Over 2500 Listed)
    Racial Slur Database - By SLUR (over 2500 listed) Slur Represents Reasons/Origins 539 Jews Corresponds with the letters J-E-W on a telephone. 925 Blacks Police Code in Suburban LA for "Suspicious Person" 7-11 Arabs Work at menial jobs like 7-11 clerks. Refers to circumcision and consumerism (never pay retail). The term is most widely used in the UK where circumcision among non-Jews or non- 10% Off Jews Muslims is more rare, but in the United States, where it is more common, it can be considered insulting to many non-Jewish males as well. 51st Stater Canadians Canada is so culturally similar to the U. S. that they are practically the 51st state 8 Mile Whites When white kids try to act ghetto or "black". From the 2002 movie "8 Mile". Stands for American Ignorance as well as Artificial Intelligence-in other words...Americans are stupid and ignorant. they think they have everything A.I. Americans and are more advanced than every other country AA Blacks African American. Could also refer to double-A batteries, which you use for a while then throw away. Abba-Dabba Arabs Used in the movie "Betrayed" by rural American hate group. ABC (1) Chinese American-Born Chinese. An Americanized Chinese person who does not understand Chinese culture. ABC (2) Australians Aboriginals use it to offend white australians, it means "Aboriginal Bum Cleaner" Means American Born Confused Desi (pronounced day-see). Used by Indians to describe American-born Indians who are confused about their ABCD Indians culture. (Desi is slang for an 'countryman').
    [Show full text]